design puzzle : avoid collision with protein where the target protein is sticked

Case number:699969-2009946
Topic:General
Opened by:nspc
Status:Open
Type:Suggestion
Opened on:Sunday, July 5, 2020 - 13:16
Last modified:Monday, July 6, 2020 - 17:26

http://pcmappeur2.free.fr/foldit/forum/collision.jpg

In design puzzle, we create a protein that have hydrogen bouds with a zone of a target protein.

This target is often stick with an other protein in real life, and we have to make sure there is no collision on it with our protein.

There is a zone that have movable sidechains in the target, it helps us to know were we have to create bouds. But it is not enough.

We can have collisions with the other sticked proteins, and we can't know that in game.

It can be interesting in some designs to have contact with target protein, not only in the movable sidechain zone, but a little around.
Because we need something that stay stick, and sometimes a part of the shape around can be used.

http://pcmappeur2.free.fr/foldit/forum/using_shape.jpg

It can be usefull to avoid mistake too. This is already hapened with hight score proteins.

To resolve this : We can just add a zone that represent a part of the protein where the target is sticked, or just add a zone that make clashs if our protein go on it.

(Sun, 07/05/2020 - 13:16  |  2 comments)


agcohn821's picture
User offline. Last seen 3 hours 13 min ago. Offline
Joined: 11/05/2019
Groups: Foldit Staff

Suggestion

Hi NSPC--thank you so much for your suggestion--I have passed it along to the team.

bkoep's picture
User offline. Last seen 1 day 12 hours ago. Offline
Joined: 11/15/2012
Groups: Foldit Staff

Great idea! (And thank you for the helpful illustrations!)

You're right: many Foldit puzzles contain only a fraction of the target protein. It might be possible to make a solution that looks good with that fragment, but would clash with the full target protein.

We've discussed a few fixes for this problem, and I like your suggestion. We may implement something like that in the future. For now, our current approach seems to be working okay. Looking at recent puzzles, very few Foldit solutions actually run into this problem.

Sitemap

Developed by: UW Center for Game Science, UW Institute for Protein Design, Northeastern University, Vanderbilt University Meiler Lab, UC Davis
Supported by: DARPA, NSF, NIH, HHMI, Amazon, Microsoft, Adobe, RosettaCommons